Commercial synthesis methods produce nanofibers of either carbon single wall nanotubes (SWNTs) or carbon multiwalled nanotubes (MWNTs) as a soot-like material. The strength and elastic modulus of individual carbon nanotubes in this soot are well known to be exceptionally high, ˜37 GPa and ˜0.64 TPa, respectively, for about 1.4 nm diameter SWNTs (R. H. Baughman, A. A. Zakhidov, and W. A. de Heer, Science 297, 787-792 (2002)). Relevant for applications needing strong, but lightweight materials, the density-normalized modulus and strength of individual SWNTs are even more impressive, being factors of ˜19 and ˜54 higher, respectively, than for high-tensile-strength steel wire.
A critical problem hindering applications of these and other nanofibers is the need for methods for assembling these nanofibers into long yarns, sheets, and shaped articles that effectively utilize the properties of the nanofibers. Since such nanofibers can confer functionalities other than mechanical properties, methods are needed for enhancing the mechanical properties of fibers made of the nanofibers without compromising these other functionalities. Important examples of these other functionalities, which combine with the mechanical functionality to make the fibers multifunctional, are electrochromism, electrical and thermal conductivity, electromechanical actuation, and electrical energy storage.
Methods are known for growing both single wall and multiwalled nanotubes as forests of parallel aligned fibers on a solid substrate and for utilizing MWNT forests for a process to produce nanofiber assemblies (K. Jiang et al., Nature 419, 801 (2002)); and in U.S. Patent Application Publication No. 20040053780 (Mar. 18, 2004). However, the resulting assemblies are extremely weak, so they cannot be used for applications that require any significant level of tensile strength.
Although advances have been made in spinning polymer solutions or polymer melts containing either SWNTs or MWNTs, the melt viscosity becomes too high for conventional melt or solution spinning when the nanotube content is much above 10%. Nevertheless, impressive mechanical properties have been obtained for polymer-solution-spun SWNTs, which in large part can be attributed to the mechanical properties of the nanotubes (see S. Kumar et al. Macromolecules 35, 9039 (2002) and T. V. Sreekumar et al., Advanced Materials 16, 58 (2004)). Another problem with both polymer melt and polymer solution spinning is that the nanotubes are not present in sufficient quantities in the polymer to effectively contribute to such properties as thermal and electrical conductivities. Additionally, the unique mechanical properties of the individual nanotubes are diluted, since by far the major component of the fiber is polymer.
A. Lobovsky et al. (U.S. Pat. No. 6,682,677) have described a sheath-core melt spinning process that attempts to avoid the usual limitations caused by low concentrations of carbon nanotubes in melt spun yarns. This process involves melt compounding 30 weight percent of very large diameter carbon MWNTs (150-200 nm in diameter and 50-100 microns in length) in a polypropylene matrix. This nanotube/polymer mixture was successfully spun as the sheath of a sheath/core polymer that contains polypropylene as the core. Despite the high viscosity of the nanotube/polymer mixture in the sheath and the brittleness of the solidified composition, the presence of the polymer core permitted this sheath-core spinning and the subsequent partial alignment of nanotubes in the sheath. Pyrolysis of the polypropylene left a nanotube yarn that is hollow (with outer diameter 0.015 inch and inner diameter 0.0084 inch). To increase the strength of the hollow nanotube yarn, it was coated with carbon using a chemical vapor deposition (CVD) process. Even after this CVD coating process, the hollow nanotube yarns had low strength and low modulus and were quite brittle (see Lobovsky et al. in U.S. Pat. No. 6,682,677).
A gel-based process enabled spinning continuous fibers of SWNT/poly(vinyl alcohol) composites (B. Vigolo et al., Science 290, 1331 (2000); R. H. Baughman, Science 290, 1310 (2000); B. Vigolo et al., Applied Physics Letters 81, 1210 (2002); A. Lobovsky, J. Matrunich, M. Kozlov, R. C. Morris, and R. H. Baughman, U.S. Pat. No. 6,682,677; and A. B. Dalton et al., Nature 423, 703 (2003)). Present problems with this process result from the fact that the nanotubes are simultaneously assembled in combination with poly(vinyl alcohol) (PVA) to form gel fibers, which are converted to solid nanotube/PVA fibers. This PVA then interferes with electrical and thermal contact between carbon nanotubes. The PVA can be removed by thermal pyrolysis, but this severely degrades the mechanical properties of the fibers.
Unfortunately, the polymer-containing fibers made by the above gel spinning processes are not useful for applications as electrodes immersed in liquid electrolytes because they swell dramatically (by 100% or more) and lose most of their dry-state modulus and strength. This process means that these polymer-containing fibers are unusable for critically important applications that use liquid electrolytes, such as in supercapacitors and in electromechanical actuators (R. H. Baughman, Science 290, 1310 (2000)).
In another process (V. A. Davis et al., U.S. Patent Application Publication No. 20030170166), SWNTs were first dispersed in 100% sulfuric acid and then wet-spun into a diethyl ether coagulation bath. Though highly electrically conductive (W. Zhou et al., Journal of Applied Physics 95, 649 (2004)), such prepared yarns have compromised properties, in part due to partial degradation of SWNTs caused by prolonged contact with sulfuric acid. This degradation, which can be partially reversed by high temperature thermal annealing in vacuum, creates a serious obstacle for practical applications. Moreover, any solution- or melt-based processing method that directly forms a polymer assembly is limited to short nanotube lengths (typically a few microns) by the viscosity increases associated with polymer dispersion and formation of globules having little nanotube orientation as a result of nanotube coiling.
Y. Li et al. (Science 304, 276 (2004)) reported that MWNT yarns could be formed directly from unoriented carbon nanotube aerogels during nanotube synthesis by CVD. While a twisted yarn was pictured, the ratio of nanotube length (˜30 μm) to yarn diameter was about unity, which means that important property enhancements due to lateral forces generated by twisting were not obtainable.
Twisting together micrometer-diameter fibers to make twisted yarns having enhanced mechanical properties is well known in the art, and has been widely practiced for thousands of years. However, no successful means has been conceived in the prior art for achieving the potential benefits of yarn twisting for nanofibers that are a thousand-fold or more smaller in diameter than for the twisted yarns of the prior art. About a hundred thousand individual nanofibers would be in the cross-section of a 5 μm diameter yarn, as compared with the 40-100 fibers in the cross-section of typical commercial wool (worsted) and cotton yarns. The challenge of assembling this enormous number of nanofibers to make a twisted yarn having useful properties as a result of a twist is enormous, and the teachings of the present invention will describe the structural features that must be achieved and how they are achieved.
Reflecting these problems with prior-art technologies of nanofiber yarns, important applications have not yet been commercially enabled, such as carbon nanotube artificial muscles (R. H. Baughman et al., Science 284, 1340 (1999) and U.S. Pat. No. 6,555,945), carbon nanotube yarn supercapacitors, structural composites involving carbon nanotubes, and electronic textiles involving strong, highly conducting nanofiber yarns,
No methods of the prior art have been developed for continuously producing strong nanotube ribbons and sheets that are free of polymer or other binding agent, although said sheets would be quite valuable for diverse applications. Carbon nanotube sheets of the prior art are usually made using variations on the ancient art of paper making, by typically week-long filtration of nanotubes dispersed in water and peeling the dried nanotubes as a layer from the filter (see A. G. Rinzler et al., Applied Physics A 67, 29 (1998) and M. Endo et al. Nature 433, 476 (2005)). Interesting variations of the filtration route provide ultra-thin nanotube sheets that are highly transparent and highly conducting (see Z. Wu et al., Science 305, 1273 (2004) and L. Hu, D. S. Hecht, G. Grüner, Nano Letters 4, 2513 (2004)). While filtration-produced sheets are normally isotropic within the sheet plane, sheets having partial nanotube alignment result from applying high magnetic fields during filtration (J. E. Fischer et al., J. Applied Phys. 93, 2157 (2003)) and mechanically rubbing nanotubes that are vertically trapped in filter pores (W. A. De Heer et al., Science 268, 845 (1995)). In other advances, nanotube sheets that are either weak or have unreported strengths have been fabricated from an un-oriented nanotube aerogel (Y. Li, I. A. Kinloch, A. H. Windle, Science 304, 276 (2004), by Langmuir-Blodgett deposition (Y. Kim et al., Jpn. J. Appl. Phys. 42, 7629 (2003)), by casting from oleum (T. V. Sreekumar et al., Chem. Mater. 15, 175 (2003)) and by spin coating (H. Ago, K. Petritsch, M. S. P. Shaffer, A. H. Windle, R. H. Friend, Adv. Mat. 11, 1281 (1999)).
For electrical device applications, nanofiber sheets are needed that combine transparency, electrical conductivity, flexibility, and strength. Applications needs include, for instance light emitting diodes (LEDs), photovoltaic cells, flat panel liquid crystal displays, “smart” windows, electrochromic camouflage, and related applications.
Eikos, Inc. developed a transparent conductive coating based on carbon nanotubes (P. J. Glatkowski and A. J. David, WO2004/052559 A2 (2004)). They used solution-based technology involving carbon single wall nanotube inks. Transparent carbon nanotube (CNT) films involving a polymeric binder were made by N. Saran et al. (Journal American Chemical Society Comm. 126, 4462-4463 (2003)) using a solution deposition method. Also, transparent SWNT electrodes have been made by A. G. Rinzler and Z. Chen (U.S. Patent Application Publication No. US2004/0197546). A. G. Rinzler noticed high transmittance of a SWNT film in both the visible range and the near infrared (NIR) range (3-5 μm) (A. G. Rinzler and Z. Chen, Transparent electrodes from single wall carbon nanotubes, US2004/0197546)).
All of these processing methods are liquid based, and none provides strong, transparent, nanofiber electrode materials or those that can be self-supporting when transparent. Also, none of these methods provides nanotube-based electrodes having useful anisotropic in-plane properties, like anisotropic electrical and thermal conductivity and the ability to polarize light.
There are reports about a successful application of a non-transparent carbon nanotube film as a counter electrode in a Gräetzel photoelectrochemical cell, which uses either liquid phase or solid phase electrolytes (see K.-H. Jung et al., Chemistry Letters 864-865 (2002); and S.-R. Jang et al., Langmuir 20, 9807-9810 (2004)). However, strong, transparent nanofiber electrodes have not been available for use in dye solar cells (DSCs), although the need for them is apparent, particularly for flexible solid-state DSCs.
Additionally, none of the above-mentioned approaches have addressed the problem of charge collection or injection from such transparent CNT coatings into organic electronic devices: organic light emitting diodes (OLEDs), optical field effect transistors (OFETs), solar cells, etc. This problem requires either very low work function (w.f.) for electron injection or high w.f. for hole injection.
Nanofibers, and in particular carbon nanofibers, are well known to be useful as electron field emission sources for flat panel displays, lamps, gas discharge tubes providing surge protection, and x-ray and microwave generators (see W. A. de Heer, A. Châtelain, D. Ugarte, Science 270, 1179 (1995); A. G. Rinzler et al., Science 269, 1550 (1995); N. S. Lee et al., Diamond and Related Materials 10, 265 (2001); Y. Saito and S. Uemura, Carbon 38, 169 (2000); R. Rosen et al., Appl. Phys. Lett. 76, 1668 (2000); and H. Sugie et al., Appl. Phys. Lett. 78, 2578 (2001)). A potential applied between a carbon nanotube-containing electrode and an anode produces high local fields as a result of the small radius of the nanofiber tip and the length of the nanofiber. These local fields cause electrons to tunnel from the nanotube tip into the vacuum. Electric fields direct the field-emitted electrons toward the anode, where a selected phosphor produces light for a flat panel display application and (for higher applied voltages) collision with a metal target produces x-rays for the x-ray tube application.
Methods are known for creating both single-wall and multiwall carbon nanotubes as forests of parallel aligned fibers on a solid substrate and for utilizing such nanotube forests as cathodes (S. Fan, Science 283, 512 (1999) and J. G. Wen et. al., Mater. Res. 16, 3246 (2001)). However, the resulting forest assemblies have various instabilities at large current loads, one such instability being the flash evaporation of catalyst and carbon, followed by spark emission of light and by the transfer of CNTs from cathode to anode, thereby destroying the cathode (R. Nanjundaswamy et. al., in Functional Carbon Nanotubes, edited by D. Carroll et al. (Mater. Res. Soc. Symp. Proc. 858E, Warrendale, Pa., 2005)). Although advances have been made in creating robust forests of oriented CNTs on glass substrates (e.g., by Motorola and Samsung), such forests are still not the best solution for the nanofiber cold cathode.
One of the most challenging issues with oriented CNT arrays is the emission non-uniformity. Due to problems with screening effects and variations in CNT structure and overall sample uniformity, only a very small fraction of the CNTs emit at any given time. Thus, unless special treatment is performed (e.g., chemical or plasma), emission from such types of CNT forest cathodes is often dominated by edge emission and hot spots (Y. Cheng, O. Zhou, C. R. Physique 4, (2003)).
Stability is the second main technical issue which remains to be solved. Two primary reasons are usually responsible for the emission instability, namely the adsorption of residual gas molecules and Joule heating of the CNTs (J.-M. Bonard, et al., Appl. Phys. Lett. 78, 2775 (2001), N. Y. Huang et al., Phys. Rev. Lett. 93, 075501 (2004)). Other methods of making cold cathodes from CNTs include formation of a composite with polymeric binder (O. Zhou et al., Acc. Chem. Res. 35, 1045 (2002)) in which CNTs are not oriented. Nevertheless, impressive emissive properties have been obtained for polymer binder/SWNT cold cathodes. The field screening effect seems not to play a crucial role in randomly oriented CNTs simply due to their statistical distribution. Also, in these types of emitters, field-induced alignment is possible that might significantly enhance field emission properties. However, the same problems that exist for oriented forests of CNTs also exist in these types of emitters.
The problem with polymer binder/CNT cathodes is that the nanotubes are not present in sufficient quantities in the polymer to effectively contribute to field electron emission and also to such properties as thermal and electrical conductivities (so that the binder is destroyed by heat and current). Additionally, the unique electrical properties of the individual nanotubes are diluted, since the major component of the cathode is by far the polymer binder. Thus, the upper level of stable field emission current is significantly reduced.
A critical problem hindering applications of these carbon nanotubes (CNT) cold cathodes is the need for methods of assembling these nanotubes into the framework of a macroscopic mounting system that is sufficiently strong and suitably shaped such that the properties of the CNTs for field emission can be effectively utilized.